11 Tex. Intell. Prop. L.J. 447

Texas Intellectual Property Law Journal Spring 2003

Articles

FROM BROADCAST TO WEBCAST: COPYRIGHT LAW AND STREAMING MEDIA

Matt Jacksona1

Copyright (c) 2003 State Bar of Texas, Intellectual Property Law Section; Matt Jackson

Table of Contents

Abstract 447

I. Introduction 448

II. Introduction to Streaming and the Copyright Issues 450

III. Musical Composition Rights 452

IV. Sound Recording Rights 454

V. Digital Performance Right in Sound Recordings Act of 1995 454

VI. Digital Millennium Copyright Act of 1998 457

VII. Ephemeral Recordings 461

VIII. CARP Report 461

A. CARP Section 114 Analysis 463

B. CARP Section 112(e) Analysis 466

IX. Librarian of Congress’ Ruling 466

X. Small Webcaster Settlement Act of 2002 469

XI. Analysis 470

XII. Conclusion 478

Abstract

This article examines the copyright issues facing streaming media and the statutory license fees established by the Librarian of Congress. The article discusses the arcane requirements for digital performances, the controversy over how ephemeral RAM copies should be treated, and the current rulemaking on the statutory *448 digital performance license. Streaming

media face complex requirements detailing what constitutes a non-interactive transmission eligible for a statutory license under section 114 of the Copyright Act. Section 114 calls for a statutory license for the digital transmissions of sound recordings (digital performances) that meet certain eligibility requirements. In addition, section 112 provides for a statutory license for the ephemeral copies made in the course of those transmissions.

A Copyright Arbitration Royalty Panel (CARP) issued a report on February 20, 2002, recommending that different rates be adopted for Internet transmissions depending on whether the transmission is a simulcast by a terrestrial broadcaster or a “free standing” transmission. The fee was 7¢ per 100 performances for simulcasts and 14¢ per 100 performances for Internet-only transmissions. The Librarian of Congress rejected the rate for Internet-only transmissions and set the rate for all webcasts at 7¢ per 100 performances (non-CPB non-commercial broadcasters who transmit webcasts will pay a rate of 2¢ per 100 performances).

An analysis of the rate reveals that the section 114 license is significantly more expensive than the corresponding license for the underlying musical work (a ratio of 3.5 to 1 compared to a ratio of 1 to 1 in other countries). The new rate favors radio stations and large, established webcast aggregators over small, independent webcasters who would be better served by a percentage-of-revenue metric. The current law’s broadcasting exemption gives over-the-air radio an unfair advantage over webcasting. That exemption should be eliminated and future CARPs should be allowed to consider public interest objectives when setting the royalty rate. However, in the long run, the current rate may give less popular recording artists an opportunity to gain attention and reduce the power of the music oligopoly. Just as important, the current rate encourages new creative uses for streaming technology.

I. Introduction

“Curtain Call for Webcasts?” “Royalty Fees Killing Most Internet Radio Stations,” “Webcasters Head to Washington in Royalty Protest,” “Webcast Royalty Proposal Draws Fire From All Sides.”1 These were just a few of the headlines written in 2002 as the Library of Congress sought to establish license fees for digital performances of sound recordings. The Librarian of Congress was caught in a no-win situation. On one side were record labels seeking to exercise a new right, on the other was a nascent industry with popular opinion on its side. At the heart of the debate was a revolutionary technology called “streaming,” which allows anyone to create their own radio station on the Internet.

The Internet is the newest medium for the distribution of audio and visual content. With each new medium comes a new fight over the scope of copyright *449 protection.2 Streaming content on the web is simply the latest battleground in the fight between content producers and distributors. Congress tried to implement a compromise; content producers (in this case the record companies) would be granted copyright protection for public performances of sound recordings on the Internet but webcasters (the distributors) would be able to take advantage of a statutory license to transmit those performances. The article details the arcane requirements for digital performances, the controversy over how ephemeral RAM copies should be treated, and the current rulemaking on the statutory digital performance license.

Section 114 of the Copyright Act calls for a statutory license for the digital transmissions of sound recordings (digital performances) that meet certain strict eligibility requirements. In addition, section 112 provides for a statutory license for the ephemeral copies made in the course of those transmissions. Webcasters face complex requirements detailing what constitutes a non-interactive transmission eligible for the statutory license under section 114.

A Copyright Arbitration Royalty Panel (CARP) issued a report on February 20, 2002 recommending that different rates be adopted for Internet transmissions depending on whether the transmission is a simulcast of a terrestrial broadcast signal or an Internet-only transmission.3 The fee was 7¢ per 100 performances for simulcasts and 14¢ per 100 performances for Internet-only transmissions. The Librarian of Congress rejected the rate for Internet-only transmissions and set the rate for all webcasts at 7¢ per 100 performances (non-CPB non-commercial broadcasters that transmit webcasts will pay a rate of 2¢ per 100 performances).4 After complaints from small webcasters that the royalty rate was too burdensome, Congress passed the Small Webcaster Settlement Act of 2002 to provide temporary relief.5

This article concludes that while the current statutory rate is high, it may be a blessing in disguise. The rate is high because it overvalues the sound recording performance right as compared to the musical composition performance right. In addition, by using a per-performance metric, the rate makes it extremely difficult for small, independent webcasters to survive. However, the high rate may ultimately *450 help recording artists and the public. First, the rate makes it possible for unknown artists to

offer a lower rate as a way to gain airplay and attention. Second, the high rate encourages webcasters to develop new, creative uses for the medium rather than relying on music as a way to attract an audience. The most significant problem with the current sound recording performance right is that it continues to favor over-the-air radio over Internet streaming.

II. Introduction to Streaming and the Copyright Issues

Streaming is distinguished from other forms of music distribution by the method and purpose of transmission. Standard music downloads consist of MP3 files or other file formats that are copied completely to the user’s hard drive for later playback. The file cannot be played until it is fully downloaded. Streaming differs from this in two important respects. First, the music is performed while the file transfer is occurring. Second, when the transfer/performance is complete, no copy of the file remains on the user’s hard drive. Consequently, the streaming song is a public performance rather than a reproduction.6 Streaming is similar to a terrestrial broadcast signal other than the method of transmission. Thus, streaming is often referred to as “webcasting.” Webcasts can take many forms. They range in length from seconds to hours. They can be performed “on demand” by selecting a program that is stored on a server or they can be continuous live or prerecorded transmissions (similar to tuning in to a radio station where the user begins listening at a random point in the programming). One final distinction that some observers make is the original source of the content. A FCC-licensed, terrestrial radio station may simulcast its programming on the Internet. This is referred to by some as a radio retransmission (RR). Programming that is only available on the Internet is referred to as an Internet-only (IO) transmission.7

Record companies, which hold the copyrights to most sound recordings, argue that digital performances of sound recordings (webcasting) reduce album sales. The concern is two-fold: (1) some users possess the technological skill to make copies of streaming songs,8 and (2) the webcasts may serve as substitutes for owning copies of particular songs. Because of the scarcity of radio frequencies, most FCC-licensed stations play a large enough variety of songs that a consumer would not consider listening to the radio to be a good substitute for owning the record to *451 listen to on demand. On the Internet, the number of webcasts is unlimited, meaning that consumers can tune into very specific programs that feature a narrow range of artists and recordings. This is more likely to serve as a substitute for owning the record, especially as wireless broadband Internet access becomes more prevalent, allowing users to listen to webcasts on portable devices. From the perspective of the copyright owners, the risk of substitution increases as the user gains more influence over the choice of songs performed. Interactive, on-demand performances pose the greatest risk to record sales while completely non-interactive and broad-based song selection poses the least significant risk.

Even before the Internet and digital media increased concerns of piracy, record labels and recording artists argued that sound recordings should enjoy a performance right. The lack of a performance right for sound recordings is an anomaly in the law.9 When records are performed on the radio, the copyright owner in the musical composition earns a performance royalty but the copyright owner in the sound recording does not. This distinction is due to the fact that when Congress made sound recordings eligible for copyright protection in 1971, it specifically excluded the public performance right.10

The Copyright Act grants copyright owners a bundle of five rights: reproduction, distribution, adaptation (derivative works), public performance, and public display.11 As will be discussed in detail in this article, sound recordings do not enjoy a standard public performance right but rather a limited right controlling public performances of certain digital audio transmissions.12 For the purpose of this article, the relevant definition of a public performance is: to transmit or otherwise communicate a performance or display of the work. . .to the public, by means of any device or process, whether the members of the public capable of receiving the performance or display receive it in the same place or in separate places and at the same time or at different times.13 Webcasting implicates both the reproduction right and the public performance right. To make matters more complex, the reproduction right is implicated in two different ways. First, the webcaster must make ephemeral copies of the songs on servers to facilitate streaming. Second, the act of streaming itself creates buffer copies of the song in a computer’s RAM. To make matters even more complex, most music recordings contain two separate copyrights: the copyright in the underlying musical composition (the song) and the copyright in the performing artist’s *452 rendering of the composition (the sound recording).14 Thus a webcaster needs to obtain permission to reproduce and perform two distinct copyrighted works: the musical composition and the sound recording.

III. Musical Composition Rights

Songwriters have long enjoyed a public performance right. Any individual or business entity that publicly performs a song must get permission from the copyright owner (songwriter or composer). Public performances include radio and television broadcasts and performances in bars, restaurants, concert halls, etc. Because countless public performances occur everyday in countless locations, songwriters have formed performing rights organizations (PROs) such as ASCAP and BMI that administer their public performance rights and collect and distribute the resulting royalties. Most businesses that make frequent use of public performances, such as broadcast stations, bars, and concert halls, obtain a “blanket license” from each PRO, which allows them to perform any song administered by that organization. It is important to note that a public performance of a song may take place as a live rendition of the song or by playing a sound recording of the song (on vinyl, cassette, compact disk, MP3 file, etc.). Public performance royalties make up a substantial share of the revenue earned by songwriters.

With the advent of webcasting, the PROs established experimental rates for permission to perform musical compositions via digital streaming on the World Wide Web (web). For example, ASCAP has established a basic rate of about 1.6% of revenues or 0.048¢ per site visit.15 Similarly, BMI’s basic fee is 1.75% of revenues or 0.012¢ per page impression.16 The PROs take the position that almost every transmission of a song is a performance, even if the purpose of that transmission is to purchase a permanent copy of the song.17

*453 Similarly, the National Music Publishers Association (NMPA) argues that every streaming transmission is also a reproduction and distribution and therefore requires a mechanical license.18 The NMPA negotiated an agreement with the RIAA stating that on-demand streaming and limited duration downloads do require a mechanical license, while streaming that qualifies for the section 114 statutory license does not.19 The Digital Media Association argued that the streaming process is the same in both instances and therefore, a mechanical license should not be required for on-demand streaming either.20 NMPA is concerned that on-demand streaming could become a substitute for purchasing a recording of the song. From NMPA’s perspective, if a person can listen to the song on-demand, it is the equivalent of owning a copy of the song and should be treated as such under the law.

In its report on the Digital Millennium Copyright Act (DMCA), the Register of Copyrights took the position that the creation of buffer copies in RAM during streaming is a fair use: The sole purpose for making the buffer copies is to permit an activity that is licensed by the copyright owner and for which the copyright owner receives a performance royalty. In essence, copyright owners appear to be seeking to be paid twice for the same activity. Additionally, it is technologically necessary to make buffer copies in order to carry out a digital performance of music over the Internet. Finally, the buffer copies exist for too short a period of time to be exploited in any way other than as a narrowly tailored means to enable the authorized performance of the work. On balance, therefore, the equities weigh heavily in favor of fair use. . . . The uncertainty of the present law potentially allows those who administer the reproduction right in musical works to prevent webcasting from taking place-- to the detriment of other copyright owners, webcasters and consumers alike-- or to extract an additional payment that is not justified by the economic value of the copies at issue.21 The Register of Copyrights recommended that Congress pass a law to exempt webcasters from liability when their streaming activities create temporary buffer copies.22

*454 While the legal delineation of rights between the NMPA and the PROs is unsettled, the agreement between the NMPA and RIAA allows webcasters who offer non-interactive streaming to transmit musical compositions after securing a blanket license for the performance rights from the PROs. They do not need to secure a mechanical license in addition to the performance license. The performance right for the sound recording is much more complex.

IV. Sound Recording Rights

Unlike the musical compositions they often contained, sound recordings were not protected by federal copyright law until 1971.23 At that time Congress declined to include a public performance right for sound recordings. Thus, radio stations pay a fee to perform the underlying song, but are not required to pay a fee to perform the sound recording of the song. Therefore, the songwriter and publishing company earn a royalty, but the recording artist and record label do not. This anomaly is

usually attributed to the strength of the broadcasting lobby when the 1971 law was passed. The distinction was justified by the argument that radio airplay benefits the record companies by promoting their sound recordings.24 Another factor was the concern among the PROs (primarily ASCAP and BMI) that Congress would direct the PROs to share their royalties with the record companies rather than forcing the broadcasters to pay an additional license fee.25 When Congress overhauled the copyright law in the 1976 Copyright Act, it still declined to grant sound recordings a performance right even though the Copyright Office had recommended including such a right.26 Meanwhile, the International Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organisations, better known as the Rome Convention, had granted a limited broadcast performance right for sound recordings in 1961.27 However, the United States is not a member of the Rome Convention and so American recording artists do not receive royalties from countries that are signatories.

V. Digital Performance Right in Sound Recordings Act of 1995

In 1995, Congress finally granted a limited performance right to sound recordings with the Digital Performance Right in Sound Recordings Act of 1995 *455 (DPRA).28 The right was limited to certain digital transmissions of sound recordings, primarily interactive and subscription transmissions. The digital performance right was later amended by the DMCA in 1998 to expand the right to include some non-subscription transmissions.29

It is no exaggeration to say that the DPRA is one of the most convoluted and unreadable laws ever passed. The law is a perfect example of interest-group policymaking that has been the hallmark of copyright legislation since the beginning of the twentieth century, as described by Jessica Litman.30 Congress simply acted as mediator between two commercial industries, the recording industry (which wanted a full performance right for sound recordings) and the existing distribution industries that publicly perform sound recordings (which wanted an exemption from any performance right). Congress responded by crafting a law that would exempt most preexisting uses of sound recordings: Notwithstanding the views of the Copyright Office and the Patent and Trademark Office that it is appropriate to create a comprehensive performance right for sound recordings, the Committee has sought to address the concerns of record producers and performers regarding the effects that new digital technology and distribution systems might have on their core business without upsetting the longstanding business and contractual relationships among record producers and performers, music composers and publishers and broadcasters that have served all of these industries well for decades. Accordingly, the Committee has chosen to create a carefully crafted and narrow performance right, applicable only to certain digital transmissions of sound recordings.31

First, the public performance right only applied to digital audio transmissions of sound recordings. Therefore, all analog audio performances were exempt as well as all performances (whether digital or analog) that do not include a transmission. “To ‘transmit’ a performance or display is to communicate it by any device or process whereby images or sounds are received beyond the place from which they are sent.”32 Nor does the performance right apply to audiovisual works (i.e., motion pictures and television programs).33 The 1995 law also created an exemption for all non-subscription, non-interactive transmissions as well as certain retransmissions of non-subscription broadcast transmissions within 150 miles of the transmitter. Transmissions to business establishments are also exempt, as are retransmissions *456 of licensed transmissions. Interactive transmissions and certain subscription transmissions do require a license.

Non-interactive subscription transmissions are eligible for a statutory license if they meet certain requirements. The subscription transmission must not exceed the “sound recording performance complement,” which basically says that no more than three songs from the same album or four songs by the same artist may be transmitted within any three hour period. The purpose of this restriction, as well as a restriction on informing listeners of when songs by particular artists would be played, was to prevent users from being able to either record songs or use the webcast as a substitute for purchasing the sound recording.34

Interactive services are not eligible for a statutory license and must negotiate directly with the record companies for the right to perform their sound recordings. An interactive service is defined in part as “one that enables a member of the public to receive a transmission of a program specially created for the recipient, or on request, a transmission of a particular sound recording. . . .”35 Record companies believed that interactive services posed the greatest threat to record sales because listeners who can listen to songs “on demand” no longer need to purchase the sound recording.

Thus, the DPRA created three categories of digital transmissions: (1) exempt transmissions, which do not require a license; (2) nonexempt transmissions that are eligible for a statutory license (primarily non-interactive subscription transmissions); and (3) nonexempt transmissions that are not eligible for a statutory license (primarily interactive transmissions).

Digital Performance Rights in Sound Recordings Act of 1995

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE *457 The 1995 Act also put restrictions on the exclusive licensing agreements between copyright owners and interactive services. The purpose of these restrictions was to prevent the major record labels, which hold the rights to 90% of the popular sound recordings in the United States, from monopolizing the market for interactive digital performances.36 This restriction would ensure that competing services could offer a similar range of songs to consumers.

The most important aspect of the 1995 law was that all non-interactive, non-subscription digital audio transmissions were exempt from the new performance right for sound recordings. Terrestrial radio stations that chose to simulcast on the Internet and Internet-only webcasters could perform sound recordings for free.37 At the time the law was written, webcasting was a nascent technology and Internet connection speeds were too slow to fully utilize the technology. By 1998, webcasting had proliferated with hundreds of radio stations and webcasters streaming music on the Internet. As Congress prepared to pass the Digital Millennium Copyright Act, the RIAA successfully lobbied to insert language to the provisions of the DPRA to close the “loophole” that prevented them from licensing non-subscription webcast performances.

VI. Digital Millennium Copyright Act of 1998

The DMCA modified the exemption in section 114(d)(1). Where the 1995 Act included one exemption for “a non-subscription transmission other than a retransmission” and a separate exemption for “a non-subscription broadcast transmission,” the DMCA eliminated the first exemption. Thus, nonbroadcast transmissions such as webcasts were no longer exempt from the performance right in sound recordings. The DMCA also extended the statutory license to cover eligible non-subscription transmissions (i.e., non-interactive webcasts) and specified conditions for three preexisting DBS/cable subscription music services (DMX, Music Choice, and DiSH nerwork) as well as two preexisting mobile satellite digital audio services (CD Radio and American Mobile Radio, now known as XM and Sirius satellite radio).

*458 All transmissions eligible for the statutory license must (1) be non-interactive; (2) not cause a receiving device to switch channels (unless the transmission is to a business establishment);38 and (3) include copyright management information contained in the sound recording.39 The five preexisting subscription services cannot exceed the sound recording performance complement or publish or announce song titles in advance.40 All other eligible transmissions (non-interactive, non-subscription transmissions and subscription transmissions made by new services) (1) cannot exceed the performance complement unless it is a third-party retransmission of a broadcast transmission (as long as the broadcast station does not regularly exceed the performance complement); (2) cannot announce or publish song titles in advance (except for announcing the names of upcoming artists within an unspecified time period); (3) cannot repeat programs of less than three hours in length, archive programs less than five hours in length, or maintain archived programs of more than five hours in length for more than two weeks; or (4) transmit the same sound recording repeatedly with the same visual information (so that the song is associated with a particular advertisement or message without the copyright owner’s consent).41 Ultimately, the purpose of these detailed provisions is to prevent users from being able to predict which artists or songs will be played at a given time so that the users cannot (1) record the transmission to avoid purchasing a copy of the phonorecord or (2) tune in the transmission as a substitute for purchasing the phonorecord.42 One unique aspect of the DMCA amendment is that in order to be eligible for the statutory license for non-subscription transmissions, a website that streams music must do so for entertainment purposes rather than simply as a promotion for particular products or services.43 Thus, Ford Motor Company would not be able to use the statutory license to create a webcasting service as background music for it corporate website. However, sites that promote or sell music-related products can still take advantage of the statutory license.44

*459 Digital Millennium Copyright Act of 1998

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE

After the DMCA was passed, broadcasters argued that simultaneous Internet streaming of radio programming was exempt from licensing under section 114(d)(1)(A), which exempts broadcast transmissions. The RIAA argued that the exemption only applied to “over-the-air” transmissions and that Internet transmissions *460 must be licensed, even if it is the identical programming and source.45 The Copyright Office issued a ruling that Internet transmissions by broadcast stations are not exempt from the licensing requirements.46 The Copyright Office reasoned that Congress did not intend to exempt webcasting when the DPRA was passed in 1995 because Congress was unaware of webcasting at that time.47 Furthermore, Congress did not change the broadcast exemption in 1998 even after specifically modifying the law to encompass webcasting. Under section 114, a broadcast transmission is defined as “a transmission made by a terrestrial broadcast station licensed as such by the Federal Communications Commission.”48 The Copyright Office agreed with the RIAA and DiMA that the word “terrestrial” is meant to limit the exemption to over-the-air broadcasts and that inclusion of the words “as such” limits the exemption to the activities that are licensed by the FCC.49 In addition, Congress repeatedly used the phrase “over-the-air” to identify the broadcasts it sought to exempt.50 The Copyright Office also noted that its interpretation made the most sense since retransmissions of broadcast signals are only exempt under certain circumstances: “[t]he Copyright Office believes that the narrowly drawn safe harbors for retransmissions of radio signals illustrate Congressional intent to distinguish between a traditional over-the-air broadcast transmission of an AM/FM radio signal and a retransmission of that signal.”51

In addition, the Copyright Office agreed with copyright owners that since the section 112 exemption does not allow broadcasters to make ephemeral recordings for Internet transmissions, it is unlikely that Congress intended to exempt those transmissions in Section 114.52 The Copyright Office also noted that it would be illogical for Congress to grant broadcasters an exemption for streaming their terrestrial signals but require other parties to pay a license fee to stream the exact same signals.53 If the broadcasters were correct in their interpretation that the exemption covers all transmissions by a broadcaster, then broadcasters could create Internet-only, exempt transmissions while third parties could not. It would also allow broadcasters to avoid the specific limits on pre-announcing songs and exceeding the *461 sound recording performance complement. Therefore, the Copyright Office ruled that only the over-the-air transmissions by broadcasters are exempt from the licensing requirements of Section 114.54

The broadcasters appealed the ruling and lost in Bonneville v. Peters.55 The court held that the ruling was reasonable and within the power of the Copyright Office to make.56 The court went even further and stated in dicta that it would have reached the same conclusion as the Copyright Office if forced to answer the question itself.57 Broadcasters have appealed the decision to the Third Circuit Court of Appeals.58

VII. Ephemeral Recordings

In addition to the performance license required by section 114, webcasting entails the making of ephemeral copies of the sound recording and the musical composition. As noted earlier, the RIAA negotiated an agreement with NMPA to license the incidental reproductions related to on-demand streaming and temporary downloads of sound recordings containing musical compositions.59 As part of the agreement, the NMPA agreed not to seek royalties for the copies of songs made during the course of webcasts that are eligible for the section 114(d)(2) statutory license.

There is still the matter of the ephemeral copy of the sound recording itself. Section 112 provides for a statutory license for eligible webcasters to make a copy of the sound recording to facilitate the licensed performance transmission.60 Only those entitled to transmit a performance under sections 114(d)(1)(C)(iv) and 114(f) can take advantage of the statutory license for ephemeral recordings. This includes transmissions to a business establishment and subscription and non-subscription services that are eligible for a statutory license under 114.

VIII. CARP Report

The provisions of sections 112 and 114 direct the Copyright Office to form Copyright Arbitration Royalty Panels (CARP) if industry groups cannot negotiate acceptable royalty rates for the respective statutory licenses.61 The CARP proceeding *462 for pre-existing subscription cable/satellite and satellite radio services uses the criteria detailed in section 801(b)(1) for establishing the statutory rate. The panel “may consider the rates and terms for comparable types of subscription digital audio transmission services and comparable circumstances under voluntary license agreements.”62 The process is repeated every five years.63 The criteria under section 801(b)(1) are as follows: (A) To maximize the availability of creative works to the public;

(B) To afford the copyright owner a fair return for his creative work and the copyright user a fair income under existing economic conditions;

(C) To reflect the relative roles of the copyright owner and the copyright user in the product made available to the public with respect to relative creative contribution, technological contribution, capital investment, cost, risk, and contribution to the opening of new markets for creative expression and media for their communication;

(D) To minimize any disruptive impact on the structure of the industries involved and on generally prevailing industry practices.64

In a significant departure from these criteria, the CARP proceeding for new subscription services and eligible non-subscription services (such as non-subscription webcasts) establishes a more limited set of criteria for establishing rates. The panel “shall establish rates and terms that most clearly represent the rates and terms that would have been negotiated in the marketplace between a willing buyer and a willing seller.”65 Thus, the panel may not consider goals A through D listed above when setting the statutory rate. The rate-setting procedure is repeated every two years.66

In February of 2002, a CARP issued its report recommending statutory rates for non-interactive, non-subscription, eligible transmissions (i.e. webcasts). The CARP distinguished between three types of webcasts targeted to the general public: (1) simultaneous retransmissions of over-the-air radio broadcasts; (2) all other Internet transmissions; (3) and transmissions by non-CPB affiliated noncommercial broadcasters.67

*463 A. CARP Section 114 Analysis

The statute requires that the panel establish “rates and terms that most clearly represent the rates and terms that would have been negotiated in the marketplace between a willing buyer and a willing seller.”68 One question was whether the RIAA should constitute a single seller (in which case the private agreements it reached with webcasters would be the appropriate benchmark) or whether the panel should envision a hypothetical market with multiple collective agencies, each able to license all sound recordings.69 The panel chose a middle ground “where the buyers are DMCA-eligible . . . services, the sellers are record companies, and the product being sold consists of blanket licenses for each record company’s repertory of sound recordings.”70

The RIAA had entered into agreements with twenty-six separate webcasters and offered those agreements as a benchmark for the statutory rate. The RIAA proposed the following rates:

RIAA Proposed Royalty Rates

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE The webcasters countered with a fee based on what radio stations pay to perform the underlying musical works.71 They then raised this amount slightly because webcasters play an average of fifteen songs per hour compared to an average of eleven songs per hour on the radio.72 They then discounted the rate based on the “promotional value” of streaming in helping record companies to sell their sound recordings.73 Ultimately the webcasters proposed a payment of 0.014¢ per song or *464 0.21¢ per hour per listener.74 In addition, webcasters argued there should be no distinction between the different types of webcasting services as long as they all meet the requirements for the statutory license, and they argued that no minimum fee should be imposed.

Applying the statutory factors, the CARP concluded that it could not determine whether webcasting had a positive or negative effect on record sales. In an important aspect of its decision, the CARP determined that the license should be based on a per-performance fee rather than set as a percentage of revenues. The panel noted that a per-performance metric is directly connected to the right being licensed and percentage of revenue is difficult to implement because of the difficulty in determining which revenues are related to the performance. Finally, a revenue metric would mean that copyright owners would not be fully compensated because many webcasters generate very little revenue.75

Ultimately the panel rejected the webcasters’ theoretical model for determining royalty rates in part because it is based upon a dozen questionable assumptions, including different rights (song versus sound recording) and different technologies (radio versus Internet). The panel then looked at the twenty-six RIAA license agreements as a potential benchmark for the statutory rate.

The panel rejected twenty-five of the twenty-six agreements as benchmarks because the agreements reflected artificially high rates to influence the CARP proceeding and many of the companies were unable to sustain operations after agreeing to the RIAA rates. Many of the webcasters who negotiated with the RIAA were in an unfair position due to ongoing litigation and negative publicity among other reasons.76 The panel felt that the RIAA had only agreed to deals that were skewed highly in its favor to influence the rate set by the panel.

The CARP did pay credence to the agreement between Yahoo! and the RIAA. Yahoo! is a webcast aggregator that streams hundreds of Internet-only (IO) channels and hundreds of broadcast radio retransmissions (RR). Because Yahoo! is a large company with significant resources, the panel felt it would be on equal footing with the RIAA during the negotiations. However, the panel noted that Yahoo!’s interests in negotiating a license were not identical to the interests of other webcasters. The panel determined that because Yahoo! was only concerned with the total license fee, it was willing to negotiate an artificially high IO rate and an artificially low RR rate in order to obtain an acceptable overall rate. The RIAA was more concerned with creating a high IO rate to influence the CARP proceeding, *465 especially since at the time of the negotiation it was unclear if radio stations would be forced to pay any license fee for their Internet transmissions.77

The CARP did accept the premise that the marketplace would establish lower rates for radio retransmissions than Internet only transmissions because of the promotional value of radio broadcasts.78 It then determined the IO rate should be the midpoint between the artificially high IO rate from the Yahoo! agreement and the “blended rate” (the overall rate Yahoo! paid when combining payments made under the IO and RR rates).79 The midpoint was 0.14¢ per sound recording per listener. The panel used the same reasoning to raise the artificially low RR rate to 0.07¢ per sound recording per listener.80 Later the panel noted that the lower rate for RR transmissions was also warranted because there was some evidence that record companies perceive less risk to record sales from typical radio retransmissions than the niche programming provided on most Internet-only transmissions (which typically focus on a narrow genre and are thus more likely to serve as a substitute for record purchases than broad-based radio playlists).81

The panel rejected the RIAA’s claim that B2B webcasters (those that provide syndicated webcasting services for third parties) should pay a higher rate. The panel also rejected the RIAA’s contention that listener-influenced webcasts should pay a higher royalty rate. The panel noted that the RIAA claimed that these services are ineligible for a statutory license to begin with. The panel did not feel equipped to define the myriad services that would qualify for the statutory license but yet should be charged a higher royalty rate: “We conclude that so long as a service complies with, and is deemed eligible for the statutory license, it should not pay a separate rate based upon listener influence.”82 The panel also reasoned that archived radio programming, and other Internet programming by radio stations that is not a simulcast of the station’s over-the-air signal, should be licensed at the IO rate (0.14¢ per sound recording per listener).

In setting a rate for noncommercial broadcasters that were not already covered by an agreement between the RIAA and NPR-affiliated stations, the panel accepted the RIAA’s argument that a prior CARP had granted non-commercial broadcasters a 2/3 discount from the commercial royalty rate for musical compositions. Therefore, the panel set a rate of 0.02¢ for simulcasts and 0.05¢ for archived *466 programming and up to two side channels (1/3 the rate for commercial RR transmissions).83

B. CARP Section 112(e) Analysis

As noted earlier, streaming requires ephemeral copies of sound recordings. Each webcaster is allowed to make one exempt ephemeral copy, but any additional copies are subject to a statutory license as provided by section 112(e). Also as noted earlier, the Copyright Office believes ephemeral copies have no independent value and that Congress should amend the law to eliminate the license requirement. In looking at the twenty-six agreements between the RIAA and webcasters, the panel again looked most closely at the Yahoo! agreement, which is this case was a flat fee that worked out to 8.8% of the total section 114 royalties. Because eight of the other agreements between the RIAA and webcasters had ephemeral license rates of approximately 10%, the panel decided to “round up” from the Yahoo! agreement of 8.8% and set the license fee for ephemeral recordings at 9.0%.84 The panel also set a minimum fee of $500 for combined section 112 and 114 licenses.

Final CARP Panel ruling on relevant webcast royalty rates

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE

IX. Librarian of Congress’ Ruling

After the CARP panel issued its report, the Register of Copyrights recommended that the Librarian of Congress reject some of the rates set by the panel.85 Predictably, the RIAA argued that the rates were set too low and the webcasters argued that the rates were set too high. The webcasters also argued that the CARP panel was arbitrary in not establishing an alternative rate based on percentage of *467 revenue. The Librarian noted that while the panel could have included an alternative fee structure, it was not obligated to do so.86 The Librarian did indeed modify the rates, most importantly by reducing the standard Internet-only (IO) rate from 0.14¢ to 0.07¢; the same rate used for radio retransmissions (simulcasts of over-the-air radio stations).

The key decision of the Librarian (based on the Register’s recommendation) was to reject the CARP panel’s conclusion that simulcasts of radio signals should enjoy a lower rate than Internet-only transmissions. The panel had based its conclusion on the fact that the RIAA had negotiated separate rates with Yahoo! for the two different types of webcasts. The Librarian determined that the panel’s conclusion was arbitrary based on the evidence: The question, however, is whether the rates in the Yahoo! agreement represent distinct valuations of Internet-only transmissions and radio retransmissions. Ultimately, the Register concludes that they do not and, therefore, the Panel’s reliance on these specific rates for IO transmissions and radio retransmissions as a tool for setting the statutory rates is arbitrary. The fundamental flaw in the Panel’s analysis, though, is not its acceptance of the Yahoo! agreement as a starting point. Rather, it is the Panel’s determination that the differential rate structure reflects a true distinction in value between Internet-only transmissions and radio retransmissions based upon the promotional value to the record companies and performers due to airplay of their music by local radio stations. The Panel reached this conclusion in spite of the fact that nothing in the record indicates that the parties considered the promotional value of radio retransmissions over the Internet when they negotiated these rates.87 The Librarian accepted the CARP conclusion that the distinction between RR and IO rates in the Yahoo! Agreement was “developed to effectuate particular objectives of the parties, distinct and apart from establishing an actual valuation of the performances.”88 The Librarian’s report noted that the CARP panel ignored its own conclusion that there was no determinate evidence that webcasting has a beneficial effect on record sales.

The Librarian then used the combined Yahoo! payments to determine the cost of the performances, yielding a “blended” rate of 0.065¢ per performance.89 In addition, Yahoo! had paid an effective rate of 0.083¢ for the first 1.5 billion transmissions.90 The Librarian then felt comfortable in choosing 0.07¢ as a middle ground between the two rates.91 In addition, the Librarian ruled that the exemption for retransmissions of radio transmissions within 150 miles of the transmitter does not *468 apply to Internet retransmissions.92 This conclusion was based on the fact that Congress did not provide a section 112 exemption for the ephemeral recording required to make such a transmission.93 Because the Librarian had lowered the rate for all IO transmissions (including archived programs and side channels) to 0.07¢, it similarly reduced the royalty rate for archived programs by non-commercial broadcasters to 0.02¢.94

In terms of the section 112 license fees, the Librarian ruled that the Panel should not have given weight to the twenty-five voluntary agreements which it had discounted in setting the section 114 rates: What causes concern, however, is the Panel’s reliance, even to a small degree, on the ephemeral rates set forth in eight of the 25 voluntary agreements it had previously repudiated. Such action is arbitrary unless the Panel can offer a clear explanation for its actions. It did not do so and, in fact, it stated that its review of the 26 licenses “reveals an inconsistent, rather than a consistent, pattern.”95 Therefore, the Librarian set the rate at 8.8%, based on the Yahoo! agreement.

Librarian’s Final Ruling on Relevant Webcast Royalty Rates

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE

Still to be determined are the record-keeping requirements that webcasters will have to implement to make sure copyright owners are properly compensated. This will be another crucial issue. In addition to providing information about the sound recording that is being performed, the proposed recordkeeping requirements require webcasters to provide information on the date, time, and time zone where the listener logged in to the webcast channel.96 These requirements include eighteen *469 data points for each song streamed. Webcasters have expressed concern that the reporting requirements will pose an undue burden on their operations.97 Since the Librarian issued his final rule, both webcasters and the RIAA have announced they will appeal the decision to the Court of Appeals for the D.C. Circuit.98 Two bills were introduced in the House of Representatives in 2002 to provide relief to small webcasters. On December 4, 2002 President Bush signed the Small Webcaster Settlement Act of 2002 (SWSA) into law.99 The SWSA provides temporary relief to noncommercial webcasters and small commercial webcasters.

X. Small Webcaster Settlement Act of 2002

The Act delays payments to be made by noncommercial webcasters until June 20, 2003 for the royalty period of October 29, 1998 to May 31, 2003.100 It also expands the definition of noncommercial webcaster to include tax-exempt entities under section 501 of the Internal Revenue Code.101 As noted earlier, noncommercial broadcasters who choose to webcast pay a reduced royalty rate. It is unclear under the SWSA whether the new tax-exempt entities that are allowed to defer payments (those that are not FCC-licensed broadcasters) would also be eligible for the reduced royalty rate. The original bill passed by the House specifically granted all noncommercial webcasters a reduced rate, but the final version enacted into law makes no specific mention of the applicable rate.102

In addition, the SWSA provides relief for small commercial webcasters for the royalty period from October 28, 1998 to December 31, 2004.103 The Act states that the privately negotiated rates agreed to under the act shall not be considered in future royalty proceedings or by the Court of Appeals in its review of the Librarian’s earlier royalty determination.104 The Act also requires the Comptroller General to submit a report to Congress on the effect of the new rates by June 1, 2004.105

*470 On December 18, the Coppyright Office issued a notice that SoundExchange (the RIAA’s royalty collection agency) and the Voice of Webcasters (a group representing small, commercial webcasters) had reached an agreement for reduced royalty rates based on a percentage of revenue and expenses metric.106 Under the agreement, small webcasters are defined as those who earned less than one million dollars in gross revenue from November 1, 1998 to June 30, 2002, and earns less than $500,000 including its affiliates in 2003, and less than $1,250,000 including third-party and subscription revenues in 2004.107 Small webcasters may elect to pay 8% of gross revenues or 5% of expenses (whichever is greater) for the period October 28, 1998 to December 31, 2002 instead of the per performance rate set by the Librarian of Congress.108 For 2003 and 2004, the rates would increase to 10% of gross revenues under $250,000 and 12% for gross revenues over $250,000 or 7% of expenses.109 The minimum royalty payment would remain $500 for 1998, but webcasters who choose to take advantage of the percentage of revenues or expenses option would be required to pay a minimum of $2,000 per year for 1999 to 2004.110 In addition, the payment would cover the section 112 ephemeral license fee, which would be considered 9% of the royalty payment.111 The agreement also contains reduced reporting requirements for eligible small webcasters. Whereas the proposed recordkeeping requirements require webcasters to report eighteen data points including the date, time, and time zone of each listener, under the new agreement small webcasters only have to report nine data points and only need keep track of aggregate tuning hours rather than information on each listener.112 It is important to remember that all the provisions of the SWSA expire at the end of 2004.

XI. Analysis

The royalty rate established by the Library of Congress works out to 1.14¢ per listener per hour for a webcaster that plays fifteen songs per hour or 0.84¢ per hour for a radio simulcast that averages eleven songs per hour.113 Noncommercial *471 stations would pay 0.24¢ per listener per hour (eleven songs) to 0.33¢ per hour (fifteen songs). Using data from Measurecast for the week of September 30 to October 6, 2002 for the top twenty webcasters, the average listener tuned in for five hours per week, with a range of two to nine hours per week.114 This means the commercial webcaster performing 15 songs per hour would pay an average fee of 2.28¢ to 10.26¢ per listener per week. A noncommercial FCC-licensed webcaster would pay 0.65¢ to 3.0¢ per listener per week for the same number of performances.115

While the RIAA has argued that its works are being undervalued, most of the protest and hyperbole has come from

webcasters. Most webcasters earn little to no revenue from their transmissions and argue they will be forced to shut down operation because they cannot afford the royalty payments.116 They feel that the statutory rate is unrealistically high. Part of the challenge in setting industry-wide rates that reflect a “willing buyer-willing seller” standard is the huge disparity between different groups of buyers. Yahoo!, RealNetworks, and other large webcasters distribute hundreds of channels and are owned by multi-billion dollar corporations. On the other side of the divide, thousands of small webcasters have almost no resources or financial backing.

Large entities like Yahoo! have three distinct advantages over small webcasters. First, Yahoo! probably has more resources to weather the initial losses involved in webcasting than an individual attempting to start a webcasting business out of her living room. Second, as a known brand, Yahoo! will probably find it easier to attract advertising revenues than a small independent webcaster even if the audience size for both entities is the same. Finally, Yahoo! may reap benefits from cross-promotion and audience spillover to its other sites.

One of the controversies that has emerged since the CARP issued its report is the admission of Mark Cuban, former president of Broadcast.com (now owned by Yahoo!), that Yahoo! was willing to agree to a per performance royalty rate in part to drive smaller webcasters out of business. In a letter to the Radio and Internet Newsletter (RAIN), Cuban wrote: I, as Broadcast.com, didn’t want percent-of-revenue pricing. Why? Because it meant every “Tom, Dick, and Harry” webcaster could come in and undercut our pricing because *472 we had revenue and they didn’t . . . . As an extension to that, I also wanted there to be an advantage to aggregators. If there was a charge per song, it’s obvious lots of webcasters couldn’t afford to stay in business on their own. THEREFORE, they would have to come to Broadcast.com to use our services . . . .117 This letter highlights the fact that the Yahoo! deal, which formed the basis of the CARP rates and the Librarian’s final rate determination, is an inappropriate benchmark and puts the thousands of small webcasters at a significant disadvantage. The fact that the RIAA was willing to agree to a lower rate for small webcasters is evidence that the recording industry also recognizes that one rate is not appropriate for all situations.118

Two other studies also suggest the current royalty rates are burdensome. A report issued by Jupiter Research concludes that under current revenue models, most webcasters cannot afford the royalty fees.119 NetRadio, would have paid more than $725,000 in royalties if it were still operating, consuming more than 40% of its revenues. A paper by Carol Ting and Steven Wildman at Michigan State University concludes that the distribution costs of Internet radio alone are higher than the realized revenues for most webcasters.120 Thus, webcasters were struggling to make a profit even before sound recording royalties were imposed. In addition, unlike over-the-air broadcasting, distribution costs for webcasters increase significantly as the size of the audience increases.121

This is not to suggest that webcasters should be allowed to perform sound recordings for free. Just as songwriters earn performance royalties, performing artists should earn royalties as well. The fact that webcasters face high distribution costs and low revenues does not by itself justify a low royalty rate.

What then should be the appropriate benchmark for the royalty rate? At the outset it is important to note that the CARP and Librarian were constrained by the specific rate-setting criteria established by Congress. Most CARPs are allowed to consider the criteria under section 801(b)(1) which allow the CARP to consider public policy objectives such as maximizing the availability of creative works and *473 minimizing the disruptive impact on the structure of the involved industries.122 However, the rates for eligible non-subscription transmissions must be “rates and terms that most clearly represent the rates and terms that would have been negotiated in the marketplace between a willing buyer and willing seller.”123 The CARP was not free to consider other policy objectives in setting the rates. Section 114(f)(2) includes two additional criteria to be used in determining the willing buyer/willing seller rates: In determining such rates and terms, the copyright arbitration royalty panel shall base its decision on economic, competitive and programming information presented by the parties, including --

(i) whether use of the service may substitute for or may promote the sales of phonorecords or otherwise may interfere with or may enhance the sound recording copyright owner’s other streams of revenue from its sound recordings; and

(ii) the relative roles of the copyright owner and the transmitting entity in the copyrighted work and the service made available to the public with respect to relative creative contribution, technological contribution, capital in-vestment, cost, and risk.124 These criteria do not add flexibility as they simply reflect the considerations that willing buyers and sellers will take

into account in the normal course of their negotiations.

It is useful to compare the royalties paid for the sound recording to the royalties paid for the underlying musical composition. In a prior proceeding for preexisting subscription services, the CARP recognized that the PRO license fees were useful in establishing a range of rates, but the fees were not determinative of the marketplace value for the sound recording performance rights.125 In the proceedings for non-subscription webcasts, the webcasters offered a theoretical model based on the license fees paid to the PROs.126 The CARP panel and the Librarian rejected that model for a number of reasons. First, the theoretical model is based upon numerous assumptions, including that the performance rights in musical works and sound recordings are analogous, different delivery systems (over-the-air vs. Internet) are analogous, and that different royalty metrics (percentage of revenue vs. per-performance) can be converted into one another.127 In addition, the *474 panel rejected the notion that the market for musical works is analogous to the market for sound recordings.128

Had the panel adopted a percentage-of-revenue model, it may have been more willing to use the PRO fees as a guide. Such was the case when a CARP set the royalties for pre-existing subscription services.129 However, since the PROs license the underlying compositions to broadcast stations on a percentage-of-revenue basis, attempting to convert that fee to a per-hour or per-listener fee is difficult. Indeed, the panel rejected the webcasters’ attempt to make such a conversion, in part because the resulting metric did not accurately predict the actual fees paid by various radio stations.130

However, both BMI and ASCAP do offer webcasters a blanket performance license for the underlying musical work based on listener hours.131 This allows us to attempt to make some comparisons. As a simple hypothetical example, assume a station averages 1000 listeners per hour. The total cost for the section 112 and 114 licenses would be $11.42 compared to a fee of 48¢ for ASCAP performance royalties for the underlying songs. The BMI fee would be similar, leading to a total cost of approximately $1.00 for the song performance licenses.132

Cost Comparisons for Licensing Schemes

Type of License Base Rate Cost for One Hour of Operation with 1000 Listeners (15 songs)

ASCAP performance license for the 1.615% of revenues or 0.048¢ per 48¢ (Note: this would be the cost even song session (up to one hour) if only one song were performed)

BMI performance license for the song 1.605% for radio simulcasts 1.75% for Unclear, but assumed to be IO webcasts or 0.012¢ per page approximately the same (or less) than impression ASCAP fee133

Section 114 license for the sound 0.07¢ per performance $10.50 recording

Section 112 license for the ephemeral 8.8% of section 114 fee 92¢ recording

*475 It is important to note that the per song rate for BMI and ASCAP increases when listeners stay tuned for less than one hour. Even if the 1000 listeners only tuned in for one song, the webcaster would pay the same PRO fees, whereas the combined section 114 and section 112 license would be reduced to 76¢ under the same conditions.

Even if we were to assume that the average listener tuned in for only twenty minutes at a time, the ratio of the sound recording license to the musical composition license remains high. The listener would hear an average of five songs in a twenty-minute span. The combined ASCAP/BMI/SESAC royalties would be approximately 0.1¢ per listener (the equivalent of 0.02¢ per song). The sound recording performance license would be 0.38¢ (0.07616¢ per song). So under the current rates, webcasters are paying more than three times as much for the sound recording performance rights than the musical composition performance rights. It is important to note that the PRO rates are experimental and may increase when the

current contracts expire.

How does this compare with other nations that recognize both rights? The Rome Convention created a sound recording performance right for over-the-air broadcasts in 1961.134 In addition, the European Union’s 1992 Copyright Directive created a similar right binding on EU members.135 An analysis of the licensing fees in twelve nations reveals that the sound recording performance fee is typically smaller than the musical composition performance fee.136 It is important to note that in each nation both license fees are based on a percentage-of-revenue. So the precise value of each right cannot easily be converted into a per-song metric. However, comparing the rates for each right can provide an indication of the relative *476 value.137 In his testimony, William Kempton concluded that the royalty rates for sound recordings “are no higher, and indeed, are generally set lower than the royalty rates for the musical composition.”138 The Librarian’s decision to set the sound recording performance rate at 0.07¢ per performance139 compared to a musical composition performance rate of approximately 0.02¢140 creates a ratio of 3.5, compared to an average ratio of 0.5 to 1.0 for the equivalent rights in other countries. This suggests that the royalty rate may be higher than the true market value of the sound recording performance right.

Two important caveats limit the strength of the preceding analysis. First, the webcasting rates established by ASCAP and BMI are experimental rates and may well increase in the future, changing the current ratio. Second, in most of the nations included in Kempton’s report, the royalty rates were set by government bodies rather than industry negotiations. Thus, the ratios used in his study do not necessarily reflect the marketplace value of the musical composition performance right and the sound recording performance right. Instead they reflect a bureaucratic or political determination regarding the relative value of each rate, just as the Librarian’s ruling is ultimately a bureaucratic rather than marketplace determination.

The panel also noted that “in determining the prices to which willing buyers and willing sellers would agree, the ‘true’ relative value--even if that could be precisely ascertained--is less important that the parties’ perception of that relative value.”141 Thus, the panel was caught in the circular logic of their mandate; to set the fee based upon what the marketplace would have determined on its own if no statutory fee were necessary. There is a particular flaw in the panel’s logic that the perceived value is more important than the true value. In a correctly functioning marketplace, the two values should be the same. Instead we are left with a situation where the last remaining property holder can hold out for a higher rate.142 The panel was apparently blind to this discrepancy even though they dealt with an identical *477 situation in evaluating the Yahoo! agreement. The Yahoo! agreement included one rate for IO transmissions and a different rate for RR transmissions. The panel noted that Yahoo! was less concerned with the value of each individual rate and more concerned with the total payments it would have to pay.143 Similarly, webcasters are less concerned with the precise song performance rate and the sound recording performance rate and are more concerned with the total royalty fees they are forced to pay. It doesn’t matter if those royalties are distributed equally between the songwriter and the recording artist or not. Since webcasters already knew the PRO fees they would have to pay to perform the song, they would be willing to pay any amount for the sound recording license fee up to the combined value of both rights. For example, if a webcaster is willing to pay up to 0.1¢ per performance and she has already agreed to pay 0.02¢ for the song performance right, then she will be willing to pay up to 0.08¢ for the sound recording performance right. This does not mean that she perceives the sound recording to be more valuable than the song. By setting the royalty fee in this manner, the Librarian is essentially preventing the PROs from raising their rates in the future, since the sound recording rate captures all of the additional value of a webcast performance.

Webcasters face an uphill battle with high distribution costs and the new royalty payments. It is unclear that webcasters can generate enough revenues to survive. Current advertising rates vary dramatically for webcasters but tend to range from a CPM (cost to reach 1,000 listeners) of $2 to $10. Live365.com’s rate card offers a variety of options; one purchase plan creates an effective CPM of $2.33.144 If the webcaster were able to sell one advertisement per song, 30% of the revenues would be allocated for the sound recording royalty. Given the nature of banner ads, it is difficult to determine the approximate inventory and revenues based on these rates. If the webcaster were able to charge a higher CPM and sell multiple ads per song, the statutory license would be a significant expense, but not necessarily insurmountable. Yet it seems clear that in the current market, such revenue projections are unrealistic.145 The Jupiter Research report concludes that the only profitable content for webcasting in the near term is news/talk and sports since these formats do not make widespread use of sound recordings.146

It is easy to sympathize with the Internet hobbyist who decides to share his or her music collection by webcasting. Many webcasters set up Internet radio stations for fun rather than profit. Economists can endlessly debate whether the specific royalty rates adopted by the Librarian are “fair” to recording artists or webcasters. However, there is something surreal about a law that requires an arbitration panel to *478 establish rates that “would have been negotiated in the marketplace.” The obvious answer would be to let the marketplace set the rates. Of course, unequal market power makes that problematic to say

the least. At the same time, to the extent that broadcasters and webcasters argue that the rates are higher than the true market value, nothing prevents them from negotiating with record companies for better rates. One record company has already made headlines by offering to waive the license fee for one year.147 This will encourage webcasters to play those sound recordings. In a truly competitive market, other record companies would soon try to get their recordings played as well (assuming they believe such performances have value).

The normal justification for compulsory licenses is the high transaction costs of dealing with countless copyright owners. Yet most webcasts focus on a much narrower range of music. In addition, while there are thousands upon thousands of different copyright owners for the underlying songs, there are a relatively small number of copyright owners who control the sound recordings. Therefore, the transaction costs are not insurmountable. A webcaster that plays folk music would only need to negotiate with labels offering that particular genre. This is one area where the market should be able to develop a clearinghouse on its own, just as there is a clearinghouse for book reproduction rights that still allows each publisher to set its own license rate. It is also important to remember that the statutory rate represents the ceiling in terms of costs. It is very likely that webcasters would be able to negotiate cheaper rates by working directly with the record companies. While it is true that this will impose certain transaction costs, most businesses deal with multiple suppliers. There is no evidence that the potential transaction costs are burdensome.

XII. Conclusion

The royalty rates established by the Librarian of Congress do not (and cannot) reflect the true market value of the rights at issue. Only the market can establish market rates. The law should be amended to allow the Librarian and future CARPs to consider other public policy goals. The easiest way to do this would be to direct the CARP to use the criteria established in section 801. The law should also be amended to require over-the-air broadcasters to pay the same royalty rate as webcasters. There is no excuse for establishing two separate rates based on the technology used to transmit the sound recordings. Quite frankly, the justification that “we have never imposed this cost before” or “we have always done it this way” is a pathetic excuse for a public policy decision. Anyone who has studied telecommunications regulation can attest to the problems caused by regulating based on technology in an age of convergence. This issue is discussed in more detail below.

From a public policy perspective, other key questions emerge:

*479 • Will the current structure and rates give over-the-air radio an unfair advantage over webcasting?

• Will the current structure and rates give webcasting aggregators an unfair advantage over small independent webcasters?

• Is there a reason we (as a society) should want to encourage the performance of sound recordings as compared to other types of content such as movies, plays, etc.?

• How will the current structure and rates affect the structure of the recording industry?

Will the current structure and rates give over-the-air radio an unfair advantage over webcasting? The answer to this question appears to be “yes.” Over-the-air radio currently enjoys two significant advantages over webcasting: portability and fixed distribution costs.148 Requiring webcasters to obtain a statutory license makes it even more difficult for them to compete with radio. More than 30% of radio listening takes place in automobiles.149 As 3G (third generation) wireless technology matures, consumers will be able to listen to high quality webcasts on portable devices. Thus, webcasting may eventually become a reasonable substitute for radio. Under the current licensing scheme, the sound recording license fee will put webcasters at a significant disadvantage since radio stations do not have to pay the license fee. This may be a moot point since the distribution costs already give radio a huge advantage over webcasting. As noted earlier, webcast distribution costs increase as the audience grows, while radio distribution costs (already quite low) do not.150 Therefore, even if radio stations are required to pay the license fee, they will still maintain a cost advantage over webcasters.

Webcasting does have one advantage over radio and that is its worldwide reach. Thus, webcasts can be narrowly tailored to reach a specific segment of the market that would be too small to target on a citywide basis. Advertisers are often willing to pay a premium to reach a more homogenous audience that fits the profile for the product being advertised. Whether this webcasting advantage over radio is enough to overcome the increased distribution and royalty costs remains to be seen, but it appears as though radio has the advantage for now--at least until portable Internet access becomes widespread.

Will the current structure and rates give webcasting aggregators an unfair advantage over small independent webcasters? As Mark Cuban noted in his letter to RAIN, the per performance royalty fee gives large aggregators with established *480 revenues a huge advantage over small webcast start up companies.151 Webcasters will have difficulty in obtaining any advertising revenues until they have attracted a significant audience. As Cuban pointed out, this will force most webcasters to turn to a large aggregator to distribute their channels unless they are able to negotiate reduced royalty rates with copyright owners. This will not necessarily harm diversity. In fact, since aggregators enjoy at least some economies of scale and allow for listeners to move easily between channels, it may actually enhance diversity and increase the likelihood of success for small webcasters.

Is there a reason we (as a society) should want to encourage the performance of sound recordings as compared to other types of content such as movies, plays, etc.? This broader social question reveals a fundamental anomaly in copyright law. While Congress has created statutory licenses for retransmissions of television programs by cable and satellite operators, this is the first instance where it has created a statutory license for the initial transmission of content.152 No other type of content can be transmitted on the Internet without the explicit permission of the copyright owner. Is there a reason why individuals should have special access to music on the Internet compared to newspaper stories, books, movies, or paintings?

This is not to suggest that music has no place in our cultural life. Music is a significant art form that has played a key role in culture since before recorded history. But recorded music is also a product and there is no reason why the producer should have to share that product any more than he or she wants. Each individual is free to make his or her own music and share it at will. Indeed, with the compulsory license for songs, anyone can record someone else’s music and distribute that as well.153

The purpose of copyright is to provide an incentive for creators to put their works into the marketplace. Copyright owners want to make their content available (albeit at a cost) in order to make a profit. If copyright owners unduly restrict access to content, consumers will choose more readily available content elsewhere.154 While many commentators have argued that consumers should be given unlimited access to content, the more important point is that copyright not prevent consumers from commenting on the content that is put into the “marketplace of ideas.”155 That does not mean the law needs to provide special access to streaming songs versus any other type of content.

*481 To suggest that webcasters should not pay a royalty to performing artists implies that they should not pay a royalty to songwriters either. I would argue that the prerecorded song is less socially valuable than the musical composition itself. Participating in the performance of songs can help to create strong communal ties in a way that merely playing prerecorded music does not. In addition, there is no compelling evidence to suggest that the marketplace is not “producing” enough performances of sound recordings, so there is no reason that such performances should be subsidized as a matter of public policy.

How will the current structure and rates affect the structure of the recording industry? If a recurring concern is that record companies control too many copyrights, then how does giving webcasters special access to their recordings reduce the continued relevance of the labels? If the Librarian of Congress keeps the royalty rate artificially low, record companies will have no effective way to engage in price discrimination. This would put lesser known artists at a disadvantage since they cannot effectively offer a lower “introductory” price. For example, if webcasters can perform popular recordings (in any genre of music) for a nominal fee, they have less incentive to perform unknown works, even if offered for free, since there is little cost savings compared to the popular recording. On the other hand, if popular recordings command a high performance fee, lesser known artists may be able to garner some airplay by offering a significantly lower fee. The worst case scenario is that popular recordings are also offered at a reduced rate in order to compete, which is what the webcasters are asking for in the first place.

Assume the opposite: that most record labels refuse to negotiate reasonable fees, thus curtailing the amount of webcasting. Perhaps record labels would prefer that webcasting not develop as an industry. If record companies make it more difficult for the public to hear their roster of bands, then would they not be disserving the bands themselves? Would that not lead some bands to decide that signing with a major label does more harm than good? Perhaps that would be the most effective way to reduce the importance of the record labels. In addition, performance royalties earned from webcasters may help many bands afford producing and distributing their own records without having to rely on a major record label.

The risk, of course, is that record labels will engage in anti-competitive conduct, favoring affiliated webcasters over independent webcasters, and more significantly, maintaining the status quo whereby over-the-air radio is the primary means

of promoting records. Clearly the oligopolistic “big five” record labels have an advantage over smaller labels in getting radio airplay, especially as radio also becomes more oligopolistic with increased consolidation. This leads us back to the first, and most significant problem: radio currently enjoys an unfair advantage over webcasting since radio does not need to pay a license fee to perform sound recordings. Correcting this discrepancy should be a top priority for Congress.

It is easy to hate the “music industry” for its commodification of such an essential art form. The dominance of the major record labels and radio groups compels us to cheer for independent labels and webcasters. But if we hate the major labels, then why should we care if webcasters have easy access to their recordings? *482 If we want radio (or webcasting) to be more than a promotional vehicle for “product,” then why not encourage webcasters to do more than play prerecorded music? An unintended consequence of the royalty rates may be that more webcasters turn to talk, debate, radio dramas, and other new creative uses for streaming.

One of the primary reasons webcasting has been embraced with such enthusiasm is that it allows for thousands of independent channels, each presumably playing a better selection of music than corporately owned, over-the-air radio stations. Are we not setting our sights too low if our ambition for webcasting is for it to be “like radio” but with better song selection? Will this truly enhance the democratic and participatory possibilities of this new technology? Perhaps in the end, we will say the Librarian’s decision is a blessing in disguise if it encourages us to think more creatively about what it is we want to say to each other using this new electronic soapbox.

Footnotes a1 College of Communications, The Pennsylvania State University

1 Headlines in order: Wash. Post, June 21, 2002, at E1; USA Today, July 22, 2002, at D.01; Wash. Post, May 8, 2002; Wash. Post, Aug. 8, 2002.

2 See, e.g., I. Trotter Hardy, Copyright Owners’ Rights and Users’ Privileges on the Internet: Computer RAM “Copies”: A Hit or a Myth? Historical Perspectives on Caching as a Microcosm of Current Copyright Concerns, 22 Dayton L. Rev. 423 (1997); Matt Jackson, One Step Forward, Two Steps Back: An Historical Analysis of Copyright Liability, 20 Cardozo Arts & Ent. L.J. 367 (2002).

3 Report of the Copyright Arbitration Royalty Panel, In re Rate Setting for Digital Performance Right in Sound Recordings and Ephemeral Recordings, Dkt. No. 2000-9 (Feb. 20, 2002) [hereinafter CARP Report].

4 Determination of Reasonable Rates and Terms for the Digital Performance of Sound Recordings and Ephemeral Recordings, 67 Fed. Reg. 45,239, 45,272 (July 8, 2002) (to be codified at 37 C.F.R. pt. 261) [hereinafter Register’s Report].

5 Small Webcaster Settlement Act of 2002, Pub. L. No. 107-321, 116 Stat. 2780 (2002).

6 However, in order to perform the song, part of the file is stored in a buffer. Theoretically, the buffer contains no more than a few seconds of the song at any one time and is continuously overwritten.

7 The distinction is only valuable if there is a legal difference in the classification of an RR transmission as opposed to an IO transmission. This became a source of much controversy when the CARP proposed two different rates, one for RR and a higher rate for IO transmissions. This is discussed in Section VII infra.

8 “Ripping” software makes it possible to make a permanent copy of streaming content onto the user’s hard drive. Most consumers do not possess this software and are therefore unable to make perfect digital copies of the streaming content.

9 David Nimmer, Ignoring the Public, Part I: On the Absurd Complexity of the Digital Audio Transmission Right, 7 UCLA Ent. L.

Rev. 189 (2000).

10 Sound Recording Act of 1971, Pub. L. 92-140, 85 Stat. 391 (1971).

11 17 U.S.C. § 106.

12 17 U.S.C. § 106(6).

13 17 U.S.C. § 101.

14 Not all sound recordings contain an underlying copyrighted song. Some musical compositions are in the public domain. Other sound recordings, such as nature recordings, do not contain musical compositions at all.

15 ASCAP Experimental License Agreement for Internet Sites & Services - Release 4.0, at http://www.ascap.org/weblicense/ascap.pdf (last visited Feb. 28, 2003). Any site visit lasting more than one hour is counted as multiple visits based on the number of hours.

16 Web Site Music Performance Agreement, at http:// bmi.com/licensing/forms/internet01.pdf (last visited Feb. 28, 2003).

17 In a joint statement, ASCAP, BMI and NMPA/Harry Fox Agency took the position that only “pure downloads” that result in a permanent, reproducible copy of the song would not be considered a performance. Joint Statement of Am. Society of Composers, Authors & Publishers (ASCAP), Broadcast Music, Inc. (BMI), The National Music Publishers’ Association (NMPA)/Harry Fox Agency (HFA on Internet Uses of Music, at http:// www.ascap.org/legislative/jointstatement.html (Nov. 2001)). Because record labels are unlikely to authorize the sale of MP3 files that do not contain copy protection, almost no download would qualify as a “pure download” and hence all downloads would be considered a performance as well. No court has yet ruled on this issue.

18 Section 115 of the Copyright Act provides for a compulsory license to make mechanical copies of songs once the song has been released as a recording. Once a recorded version of a song is published, anyone can distribute their own recording of the song to the public as long as they pay the statutory rate. The compulsory license does NOT extend to making copies of the original recording of the song, nor does it extend to making copies of the lyrics or sheet music. 17 U.S.C. § 115.

19 Agreement between NMPA, Harry Fox Agency and the RIAA, Section 8.1(a), at http://nmpa.org/pr/FinalRIAAAgreement.pdf (Oct. 9, 2001).

20 Comments of the Digital Media Association, at http:// www.copyright.gov/carp/dpd/dpd001.pdf (Feb. 6, 2002).

21 Register of Copyrights, U.S. Copyright Office, DMCA Section 104 Report, xxv-xxvi (Aug. 2001).

22 Id. at 142-43.

23 Sound Recording Act of 1971, Pub. L. 92-140, 85 Stat. 391 (codified as amended in scattered sections of 17 U.S.C.).

24 Kimberly L. Kraft, The Webcasting Music Revolution Is Ready to Begin, as Soon as We Figure Out the Copyright Law: The Story of the Music Industry at War with Itself, 24 Hastings Comm. & Ent. L.J. 1, 6 (2001).

25 Id.

26 Nimmer, supra note 9, at 190.

27 International Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organisations (Rome Convention), at http:// www.wipo.int/clea/docs/en/wo/wo024en.htm (Oct. 26, 1961).

28 Pub. L. No. 104-39, 109 Stat. 336 (1995) (codified as amended in scattered sections of 17 U.S.C.).

29 Pub. L. No. 105-304, 112 Stat. 2860 (1998) (codified as amended in scattered sections of 17 U.S.C.).

30 Jessica Litman, Copyright Legislation and Technological Change, 68 Or. L. Rev. 275 (1989).

31 S. Rep. No. 104-128, at 13 (1995).

32 17 U.S.C. § 101.

33 This creates an interesting loophole whereby an audio-only webcast must obtain a license but a music video webcast does not have to obtain a license. Of course, the webcaster would need to obtain a license to perform the music video but none of those royalties would necessarily be allocated to the performers.

34 S. Rep. No. 104-128, at 24-25 (1995).

35 17 U.S.C. § 114(j)(7).

36 S. Rep. No. 104-128, at 25 (1995).

37 17 U.S.C. § 114(d)(1)(B)(i-iii).

38 The purpose of this clause is to make sure webcasters cannot exceed the sound recording performance complement by switching listeners from one channel to another. A large webcast aggregator otherwise could conceivably program dozens of channels in such a way that a listener could hear an entire album by an artist by being switched from one channel to another. The exception for businesses makes little sense since transmissions made to a business establishment are apparently exempt from licensing under section 114(d)(1)(C)(iv).

39 17 U.S.C. § 114 (d)(2)(A)(i-iii).

40 17 U.S.C. § 114 (d)(2)(B)(i-ii).

41 17 U.S.C. § 114 (d)(2)(C)(i-iv). The remaining five conditions § 114 (d)(2)(C)(v-ix) simply further attempt to prevent users from being able to use these transmissions as a substitute for purchase of the sound recordings.

42 Jane C. Ginsburg, Copyright Legislation for the “Digital Millennium”, 23 Colum.-VLA J.L. & Arts 137, 167 (1999).

43 17 U.S.C. §114(j)(6) (2000).

44 H.R. Conf. Rep. No. 105-796, at 87 (1998).

45 Public Performance of Sound Recordings: Definition of a Service, 65 Fed. Reg. 77,292 (Dec. 11, 2000).

46 Id.

47 Id. at 77,301.

48 17 U.S.C. § 114(j)(3) (emphasis added).

49 Public Performance of Sound Recordings: Definition of a Service, 65 Fed. Reg. at 77,298 (Dec. 11, 2000).

50 Id.

51 Id. at 77,299.

52 Id. at 77,300.

53 Id.

54 Id.

55 Bonneville Int’l Corp. v. Peters, 153 F. Supp. 2d 763, 59 U.S.P.Q.2d (BNA) 1622 (E.D. Pa. 2001).

56 Id. at 779, 59 U.S.P.Q.2d at 1634.

57 Id. at 784, 59 U.S.P.Q.2d at 1638.

58 Radio Stations Appeal Ruling on Web Fees, N.Y. Times, July 17, 2002 at C4.

59 Agreement between NMPA, Harry Fox Agency and the RIAA, (Oct. 9, 2001), at http://nmpa.org/pr/FinalRIAAAgreement.pdf.

60 17 U.S.C. § 112(e).

61 17 U.S.C. § 112(e)(4); 17 U.S.C. § 114(f)(1)(B).

62 17 U.S.C. § 114(f)(1)(B).

63 17 U.S.C. § 114(f)(1)(C)(ii)(II).

64 17 U.S.C. § 801(b)(1).

65 17 U.S.C. § 114(f)(2)(B).

66 17 U.S.C. § 114(f)(2)(C)(ii)(II).

67 CPB-affiliated broadcasters (i.e., NPR-affiliated stations) had already negotiated a private license agreement with the RIAA. Register’s Report, supra note 4, 67 Fed. Reg. 45,239, 45,258 (July 8, 2002).

68 17 U.S.C. § 114(f)(2)(B).

69 CARP Report, supra note 3, at 22.

70 Id. at 24.

71 This required taking the yearly fee for a sample of music-intensive radio stations and calculating an average fee per listening hour and then dividing that figure by the average number of songs played per hour to obtain a per-song fee. CARP Report, supra note 3, at 28.

72 Id. at 29.

73 Id. at 30.

74 Id.

75 Id. at 37.

76 Id. at 59-60.

77 Id. at 65-69. See discussion, supra notes 45-58 and accompanying text.

78 CARP Report, supra note3, at 74-75.

79 Id. at 76. Analyzing this section of the CARP report is extremely difficult since much of the material has been redacted from the public version.

80 Id. at 77.

81 Id. at 84.

82 Id. at 81.

83 Id. at 93.

84 CARP Report, supra note 3, at 104. However, the eight agreements were not used in setting the section 114 royalty rate because of the unfair bargaining position of the RIAA.

85 Register’s Report, supra note 4, at 45,272..

86 Id. at 45,250.

87 Id. at 45,252.

88 Id.

89 Id. at 45,255.

90 Id.

91 Register’s Report, supra note 4, at 45,255.

92 Id. at 45,256. Section 114(d)(1)(B) provides the exemption within 150 miles of the transmitter.

93 Id.

94 Id. at 45,259.

95 Id. at 45,262.

96 Notice and Recordkeeping for Use of Sound Recordings Under Statutory License, 67 Fed. Reg. 5,761, 5,766 (Feb. 7, 2002) (to be codified at 37 C.F.R. pt. 201).

97 Paul Maloney, Streaming Report Requirements Seem to Follow RIAA Wishes, RAIN: Radio and Internet Newsletter (Feb. 19, 2002), at http:// www.kurthanson.com/archive/news/021902/index.asp.

98 Amy Harmon, Technology Briefing Internet: Royalty Rates Will Be Appealed, N.Y. Times, Aug. 8, 2002, at C5.

99 Pub. L. No. 107-321, 116 Stat. 2780 (2002).

100 Id. at §3(a).

101 Id. at §4.

102 Compare H.R. 5469 as passed by the House on October 7, 2002, Cong. Rec. H7043, H7045 with the final version signed into law, Pub. L. No. 107-321, 116 Stat. 2780 (2002).

103 Pub. L. No. 107-321, §4, 116 Stat. 2780 (2002).

104 Id.

105 Id. at §6.

106 Notification of Agreement Under the Small Webcaster Settlement Act of 2002, 67 Fed. Reg. 78,510 (Dec. 24, 2002).

107 Id. at 78,513.

108 Id.

109 Id.

110 Id. at 78.511. These are the basic figures for simplicity. The bill contains specific provisions regarding what counts as revenues and expenses and slightly higher minimum fees for 2003 and 2004 based on a revenue threshold.

111 Id.

112 Id. at 78,512.

113 The overall rate is 0.07616¢, and includes the section 114 rate of 0.07¢ PLUS the section 112 rate of 8.8%. The formula for calculating royalties is n x 0.07616, where n = number of performances. For noncommercial webcasters, the overall rate is 0.02176¢ per performance (0.02 + (0.02 x 8.8%)).

114 Audience turnover varies widely for each webcast channel. For the top 20 webcasters that week, Music Match’s Top 40 channel had the shortest average listening time (1.9 hours per week) and WFXZ’s classic rock simulcast had the longest average listening time (9.1 hours per week).

115 Under the Librarian’s ruling, only noncommercial broadcasters licensed by the FCC (and not part of the private NPR agreement) are eligible for the reduced rate. The Small Webcaster Settlement Act expanded the definition of noncommercial webcaster to include any webcaster that is tax-exempt under section 501 of the Internal Revenue Code. However, the terms in that act only apply to webcast performances made between October 28, 1998 and December 31, 2004 and it is not clear if the reduced royalty rate applies to tax-exempt organizations that do not hold an FCC license. See supra notes 100-102 and accompanying text.

116 Christopher Stern, Curtain Call for Webcasts? Wash. Post, June 21, 2002, at E1; Jefferson Graham, Royalty Fees Killing Most Web-Based Radio Stations, USA Today, July 22, 2002.

117 Mark Cuban, e-mail letter to RAIN: Radio and Internet Newsletter (June 24, 2002), http://www.kurthanson.com/archive/news/062402/index.asp. Note: Cuban had left Broadcast.com before Yahoo! reached its final agreement with the RIAA.

118 The RIAA agreed to the compromise in its support for H.R. 5469. When the bill stalled in the Senate, Soundexchange, the entity set up by the RIAA to collect royalty payments, agreed to defer collection of royalties until final action is taken on the bill. Jon Healey, Webcasters Saved From High Royalties, L.A. Times, Oct. 19, 2002, Business, at 1.

119 Internet Radio: Post-CARP Survival Scenarios, MUS02-U06, Jupiter Research (2002) (executive summary on file with author).

120 Carol Ting & Steven S. Wildman, The Economics of Internet Radio 14-16 (Sept. 29, 2002) (paper presented at the Telecommunications Policy Research Conference), at http:// intel.si.umich.edu/tprc/papers/2002/89/internetradio.pdf.

121 Id.

122 17 U.S.C. § 801(b)(1).

123 17. U.S.C. § 114(f)(2)(B).

124 Id.

125 Determination of Reasonable Rates and Terms for the Digital Performance of Sound Recordings, 63 Fed. Reg. 25,394, 25,404 (May 8, 1998) (to be codified at 37 C.F.R. pt. 260) [hereinafter CARP DSTRA Report].

126 CARP Report, supra note 3 at 27.

127 Id. at 39-40.

128 Id. at 41.

129 CARP DSTRA Report, supra note 125, at 25,409-10.

130 CARP Report supra note 3, at 42.

131 See ASCAP Experimental License, supra note 15 and BMI’s Web Site Performance Agreement, supra note 16.

132 SESAC, the third PRO, uses a fee structure based on number of page requests, but with a maximum fee of less than $2000 per year. ASCAP historically has charged the highest rates of any of the three PROs. While BMI was unwilling to provide precise numbers, it is doubtful that they would be significantly higher than the fees charged by ASCAP. In comparing their webcasting fees based on revenues, BMI currently charges 0.135% more than ASCAP (1.75% compared to 1.615%).

133 When contacted by the author, BMI was unwilling to clarify the page impression metric. Theoretically, the fee is based on the number of separate pages visited. Thus, if a user simply went to one page to select a webcast, that user might be able to listen to hundreds of songs but only register ONE page impression at 0.012¢. For simplicity, this article doubles the ASCAP fee and then rounds up to estimate the total song performance license fee.

134 International Convention for the Protection of Performers, Producers of Phonograms and Broadcasting Organisations (Rome Convention), at http://www.wipo.int/clea/docs/en/wo/wo024en.htm (Oct. 26, 1961).

135 Council Directive No. 92/100, Nov. 19, 1992 (European Commission).

136 Digital Performance Right in Sound Recordings and Ephemeral Recordings, Dkt. No. 2000-9, CARP DTRA 1&2, testimony of Paul William Kempton, Media Matrix Grp., at http:// www.digmedia.org/webcasting/CARP/KemptonTestimony.DOC.

137 Id. at P 11.

138 Id. at P 10.

139 This does not include the 8.8% additional fee for the section 112 license.

140 As noted above, the PROs use a “session” metric rather than a performance metric. Thus, the rate per song varies based on how long the listener tunes into the webcast. ASCAP’s fee of 0.048¢ per hour translates into a maximum of 0.048¢ per song (if the user only listens to one song) or as little as 0.0032¢ per song if the user listens to fifteen songs within a one-hour time span. The 0.02¢ is based on a twenty-minute transmission in which the user listens to five songs and includes payments to ASCAP, BMI, and SESAC.

141 Id. at 41.

142 If a property developer were buying a group of houses, the last seller can try to hold out for more money. The developer is likely to pay since that house represents the last hurdle before developing the property. Similarly, webcasters need the rights to both the song and the sound recording. Because the royalty fees for the song have already been agreed upon, the RIAA has improved negotiating power as the last seller.

143 CARP Report, supra note 3, at 64-66.

144 Live365.com Sales Presentation (on file with author).

145 See, e.g., Emma Cowing, Poor Performance of Internet Adverts Creates Buyer’s Market, The Scotsman, Dec. 4, 2001, at 5.

146 Internet Radio, supra note 119.

147 Artemis Records to Waive Webcast Fees, L.A. Times, July 30, 2002, at C3.

148 Fixed in the sense that once a radio station pays the expenses involved in transmitting its signal, the marginal cost to add an additional listener is zero. If there are only a few listeners, the relative cost per listener would be quite high, indeed higher that the costs for a webcaster. However, the situation quickly reverses as more listeners are added.

149 Arbitron, 2001 Radio Today: How America Listens To Radio 11 (2002).

150 See Ting & Wildman, supra note 120, at 6-7.

151 See Cuban, supra note 117.

152 The Copyright Act provides for compulsory licensing of cable retransmissions (17 U.S.C. § 111) and satellite retransmissions (17 U.S.C. §§ 119 and 122).

153 17 U.S.C. § 115.

154 For example, the motion picture industry restricts access to its content, and consumers end up spending more time watching television, and less time in movie theaters.

155 See, e.g., Matt Jackson, Using Technology to Circumvent the Law: The DMCA’s Push To Privatize Copyright, 23 Hastings Comm. & Ent. L.J. 607, 615-19 (2001).

11 TXIPLJ 447